Thank you for the source! I agree that Tesla should’ve halted FSD and went all in on ensuring FSD sees and stops for busses, this is dangerous and mishandled.

My original claim wasn’t that FSD or Tesla is perfect, just already safer than humans.

Please note my below sources are from 2024 to get a full year and have data before the Trump administration. This means that the software is older and so doesn’t hit my 10x claim but does hit 5x. I’m confident 2025 will hit 10x. (And if we keep to my original claim being FSD specific, ie not including autopilot it is already at 10x.)

https://www.nhtsa.gov/press-releases/nhtsa-estimates-39345-traffic-fatalities-2024

1.2 fatalities per 100 million miles

Tesla 2024 fleet size(yes a Tesla source but tesladeaths is not up to date on fleet size)

https://x.com/tesla_ai/status/1905644814483251709

2.16 billion miles 2024 Expected deaths 34

Actual deaths 6 (once again being favorable to the opposing view including autopilot deaths in this total. For comparison, even Tesla deaths.com only claims 2 fatalities involving the use of FSD use over all time)

https://www.tesladeaths.com/

(Would’ve used this source for fleet size miles as well but not updated since 2019)

I’ve used a basically entirely biased against Tesla source for deaths. With their numbers Tesla fsd is still multiple times safer than a human.

For the record, I could find no record of FSD actually hitting a child getting on a bus. It’s still a supervised system and human drivers are aware it doesn’t work with busses and take action to stop for the bus. The coming update does in fact now stop for busses so it is being addressed, albeit I agree not with the speed and seriousness such a miss should merit.

I’d rather have a system with a human and computer ensuring safety. Even if at the moment there are still for sure situations where the computer is far worse than a human, it being better in the majority of cases still makes it more safe to have than not have.